1. Reflections on Technology's Nature and Social Impact

• The interview opens with the message that we should pay more attention to technology's impact than to tech companies themselves.

  "We tend to over-glorify tech companies as a society. What really matters is the impact that technology brings."

• He mentions a longstanding desire to see real change in education, suggesting AI can drive that transformation.

  "For decades, everyone has been saying, 'Could there be a real technological intervention that brings genuine change to education?'"

2. Building New 'Scaffolding' for the AI Era

• He emphasizes that Microsoft is building infrastructure and platforms for the AI era.

  "We are definitely trying to build scaffolding for the AI era."

• He explains that we have entered a platform transition, where what matters now is not a single app but the scalability and composability of the entire platform.

  "Now it's not just about a single app — it's about how the entire platform scales and composes."

• He cites the example of Stanford Medicine's AI use cases (such as tumor board meetings):

  "Pathology, multiple labs, PubMed — all that data needs to be orchestrated by multiple agents."

• He stresses the importance of building a stack where every layer is open and composable.

  "We need to build a true stack where every layer is open, composable, and where standards and protocols exist."

3. Agent-Centric New User Experiences (UI) and Ways of Working

• He uses a metaphor to describe a world where every knowledge worker becomes an 'agent manager'.

  "A world where knowledge workers are not just knowledge workers, but agent managers — that's exactly why we're building scaffolding for the AI era."

• He draws on the history of integrated UIs like Outlook and Teams:

  "In the old days, calendar, contacts, and email were all separate — Outlook unified them. Teams did the same thing."

• He predicts that going forward, chat, search, agents, notebooks, and more will all be integrated into a single interface.

  "The goal is to bring everything — chat, search, agents, notebooks — into one interface."

4. The Changing Nature of Knowledge Work and AI Tool Adoption

• He explains that AI tools are raising the level of abstraction in knowledge work and completely flipping workflows.

  "Now even I, as CEO, type in a prompt directly to get what I need in one shot and share it with colleagues. The workflow has been completely flipped."

• He advises that it's important to actively use AI tools and personally transform your work outputs and workflows.

  "Use the tools, change your work. You have to be the agent of change in how you work."

• He acknowledges concerns about job displacement, but adds:

  "There will certainly be job displacement. But the best defense against that is skilling and reskilling."

• He emphasizes that using the tools is itself the beginning of change.

  "Using the tools — that is the start of change."

5. Innovation in Developer Work and the Future of AI Code Generation

• He describes the specific ways AI is transforming developer work:

  "30% of Microsoft's new code is already being generated by AI."

• He gives real examples of AI tool usage: code autocomplete, code explanation, flowchart generation, multi-file editing, and more.

  "Yesterday I highlighted some code and said 'draw this as a flowchart' — and it immediately visualized it."

• He stresses that collaboration between AI agents and humans is essential:

  "Ultimately, humans are in the loop. We tend to overestimate the autonomy of AI."

• He foresees a structure where AI must always pass human review before code review, and teams will work alongside AI agents within their development organizations.

  "There's always a human review before CI/CD. Ultimately, you'll be working with AI agents inside your development organization."

6. Enterprise Competitive Edge: Domain-Specific Agents and Data Utilization

• He explains that fine-tuning AI on a company's own data can be a major source of competitive advantage.

  "Tuning your Copilot system with your organization's knowledge and data — that's where true competitive advantage lies."

• He stresses that sustainable competitive advantage comes from real-world market feedback (reward functions) and iterative learning.

  "The reward function from the market — actual user feedback — is what creates a virtuous cycle."

7. Organizational Change, Culture, and Building Capability

• He emphasizes that organizations must fundamentally change their structure and ways of working to keep pace with technological change.

  "The three things we need to change are 'how we work,' 'what we do,' and 'how we go to market.'"

• Drawing on Microsoft's own history of transformation:

  "Even Windows eventually stopped growing. That's why we had to build other businesses."

• He reiterates that organizational culture and capability building are the heart of change:

  "In the end, what matters is culture and capability building. That's what allows you to seize more opportunities."

• He cautions that there are limits to learning simply by watching others succeed:

  "Someone once said, 'Watching someone else go to the gym doesn't make you healthy.' You have to do it yourself."

8. Grassroots Tool Diffusion and Tailored Learning

• He explains that the spread of tools is itself the catalyst for change:

  "When PCs spread through offices, the way people worked changed. Even without formal training, change happened naturally as the tools spread."

• He shares a real internal example from Microsoft:

  "A networking engineer who used to do things manually built their own agents and orchestrators to automate it. Anyone can do it with low-code and no-code tools."

• He mentions that Excel is the most widely used programming tool in the world:

  "We all learned programming through Excel. This kind of change is going to happen again."

9. Proactive Agents and Transparency

• He describes examples and expectations for proactive agents:

  "When the technology is powerful enough, it will feel like it disappears. The goal is to interpret user intent and handle tasks with minimal friction."

• He stresses that users must retain control and be able to transparently view an agent's work history:

  "I always check the session log. I can transparently see how every commit was made."

10. The AGI Debate and the True Value of AI

• He shares a critical perspective on the AGI (Artificial General Intelligence) debate:

  "AGI is just word games. The real value of AI is in global economic growth."

• He emphasizes that AI must deliver tangible productivity gains across society — in healthcare, education, and more:

  "19–20% of US GDP goes to healthcare. If AI can reduce that inefficiency, the impact would be enormous." "For decades we wanted 'a real technological intervention for education' — and now there is statistical evidence that tools like Copilot are actually making that change."

• He insists that it is not technology itself, but the 'use' and 'impact' of technology that deserves genuine praise:

  "It's not the tech industry itself that should be celebrated — it's the moment when technology is actually used and changes the world. That's the future we're really after."

11. Closing

• Wrapping up the interview:

  "Thank you so much. It's an honor to be part of this conversation."

• He leaves us with deep insight into the nature of technology, the changes of the AI era, and the role of human beings.

Key Keyword Summary

• Impact of technology
• Scaffolding for the AI era
• Platform transition
• Agent-centric UI
• The changing nature of knowledge work
• AI and human collaboration
• Domain-specific agents
• Organizational culture and capability
• Tool diffusion and tailored learning
• Proactive agents and transparency
• The AGI debate and the true value of AI
• Technology use and social impact

"The moment technology truly changes the world — that is the moment we should really be celebrating." "Use the tools, lead the change. That is the path to becoming a protagonist in the AI era." "Watching someone else go to the gym doesn't make you healthy. You have to do it yourself." "The true value of AI is in making all of our lives better."

😊 I hope this summary helps you understand the changes of the AI era and their true essence!